Bagging Integrated Classification Model For Credit Card Anti-fraud

Mediated by credit card financial science and technology is becoming an irreversible trend,showing an emerging trend,but with the risk of credit card fraud also escalating.Credit card fraud causes huge losses to consumers and financial companies every year.Scammers are constantly trying to find new rules and strategies to implement illegal behaviors.Building accurate and effective credit card fraud transaction identification systems has become an eternal theme in the financial field.Based on the knowledge of credit card fraud theory and the existing data mining methods such as Naive Bayes,Support Vector Machine,and K-Nearest Neighbor Algorithm,this paper proposes a credit card fraud detection model based on SmoteBagging integrated classification.The Smote algorithm is to balance the number of positive and negative samples in the training data set,so as to prevent the uneven proportion of positive and negative samples from causing significant interference to the classification;the Bagging algorithm is to assign different weights to the classification results of multiple classifiers,and combine these weighted classification results to get A more accurate classification result.75% samples of pre-processed credit card fraud data set were used as training set to train and optimize the anti-fraud model of credit card,The remaining 25% of the samples were test sets,and the effectiveness and practicability of the credit anti-fraud model were evaluated through the evaluation indicators such as accuracy,recall,F-value and G-mean,ROC curve and AUC in the confusion matrix.Empirical analysis results show that the Smote-Bagging classification model has high recognition accuracy in the detection of credit card fraud.Its model and algorithm deployed in the financial system can effectively identify credit card fraudulent consumption and minimize the losses of consumers and financial institutions.